Quote Originally Posted by Eri523 View Post
I've never seen this in my own debugging sessions, but I'd interpret that as an address range which has not been assigned any physical memory at all (neither RAM nor page file space). I'm not sure if that's actually what you meant by "unpaged", but I think programs like Process Explorer woud call that "uncommited". This assumption is supported by the fact that the address where the "real" data starts (0x0769E000) is on a 4k boundary, which is the most commonly used memory page size. I don't think that memory has just been paged out (i.e. stored in the page file and then released the physical RAM), because then it would certainly have been paged back in transparently on demand from the debugger.

Do you get a 0xC000005 thrown for that pointer? Accessing uncommitted addresses is one of the most common reasons for this exception, even when just attempting a read.
Thanks, Eri. Yes, I did get a 0xC000005 exception.

It turns out the pointer was pointing to deallocated memory. The memory window shows 0xFEEEFEEE repeatedly (special value for deallocated heap memory) just after the deallocation, but later just before the dangling pointer is accessed, the memory window shows the question marks. Perhaps the memory page has been returned to the OS by then.

If anyone knows for sure what the question marks mean (or can give a reference), I'm still interested.